Fmam calculator

[Brian Dayton]

here's a prototype (soon to be finished based on feedback) calculator for Fmam.

In general, Fmam calculations for insulated cavities give results that are 0.7 times that of the empty cavities.  (see IR-586 for a quick reference)

as such, if one wanted to calculate Fmam for various things from IR-761, and wanted to do so based upon the traditional methods, you'd have to consider partially filled (like that oddball wall with 40mm of something or other) cavities as hybrids between the insulation calc and the empty calc.

this sheet offers these:

1.  sheet with limited selectable menu with pre-loaded masses  ( i might expand this to cover everything)
2.  sheet where mass is entered manually
3.  traditional calc sheet with manually entered mass

4.  1&2 compensate for stud volume, adjusting the volume of insulation and air


kinda fun

Brian

[Brian Dayton]

revised, updated, a little nicer, with some comments on the relevant background theory and what not

deleting other file now, am done improving it unless error is found


for likers of modeling software:  the oft-quoted modeling software, insul, calculates Fmam based upon empty cavity in the trial version.  you may use this spreadsheet to calculate the equivalent cavity depth for the insulation you want to model, then enter that into insul.  

that'll help a bit.

that doesn't mean i condone software as a LF modeling tool

[Bob]

Brian:

the oft-quoted modeling software, insul, calculates Fmam based upon empty cavity in the trial version

My demo version of insul48SA.exe says it uses fiberglass. Hmm, it's commented out, but it's there in the picture on top left.

For a double stud wall, two layers of gypsum 5/8" on both sides, 2x4 studs, 16" centers, 3.5" insulation, yours says 40hz. Insul on the other hand predicts 40hz for that wall.

[Brian Dayton]

bob

the demo version of insul says it uses some insulation or other, yes, but the calculation for Fmam is in-line with the assumption that no insulation is present.

on the full-version (which i've seen), Fmam is in-line with full-insulation from my spreadsheet, and also TL varies a bit depending on which insulation you select (At higher frequencies) and how thick it is.

so what insul calculates is Fmam, no compensation for insulation (demo version)

to compensate for insulation, calculate the equivalent cavity depth via this sheet, and plug it in

this will affect mid/high frequencies due to the transition from pressure zone to wave field (that midrange dip in double stud walls) that insul incorprates

[Brian Dayton]

it is a happy day in brianspointlessmentalfascinationsville

i found an article i hadn't seen before, see here:

http://irc.nrc-cnrc.gc.ca/fulltext/nrcc ... c44692.pdf


They model the behavior of resilient channel walls based upon (get this ;) )

1.  stiffness of the air cavity
2.  stiffness of the resilient channel

and outline a new calculation system that includes an equivalent air space to account for channel stiffness.  They note that the traditional calculations (air space only) are 2/3 octave or more off, but their new model is typically within a few hertz.

well, glory be.  onward and upward goes humanity, what a species.

I wonder if, in time, enough work here and there and everywhere will come into my viewport to justify have NOT done all the work on understanding low frequency wall mechanics that we did here in 2004?  Perhaps somewhere out there, each and every aspect of wall behavior that we (a group of chemical-numbed nitwits not hardly qualified to do acoustics research) sat down and fretted and pondered and built and tested and tried again...

every pea-brained, somewhat competent, and truly clever method of modeling these things that we came up with, always winding up with piles of empirical data and guesses as we lack the mathematical background to really make models in the traditional academic sense...

i wonder if all this is already known, or will soon enough be known by someone out there somewhere?  and if we could have just found all hte documents, we could have saved ourselves all the trouble?

What amazing fun reading that article was.  Also, the Lin and Garrelick article... my goodness what a thrill it was to find that article, which less or more stated everything we already knew, but stated by them 10,000 times more formally (and competently, i'm sure).


life is such a strange, strange thing.  And i really, really don't care if anybody ever reads a bit of what we did here to model walls, as long as over time it's proven to be right by those best equipped to prove it, i'll drink and be merry.

check out that article, it's awesome.  i'm gonna send the NRC a valentine next year.

[andrebrito]

Hello Brian !

How did you find that link ? It's very interesting indeed.

The Lin and Garrelick article what adress was that ?

Well I believe academically speaking lots of stuff have been studied but that information is not really acessible (and most imes not really  understandable) for the average citizen lol !  

Also in terms of models stuff like FEM and BEM and other methods of software prediction are used for more complicated matters.

Andre

[Brian Dayton]

Hello Brian !

How did you find that link ? It's very interesting indeed.

The Lin and Garrelick article what adress was that ?

Well I believe academically speaking lots of stuff have been studied but that information is not really acessible (and most imes not really  understandable) for the average citizen lol !  

Also in terms of models stuff like FEM and BEM and other methods of software prediction are used for more complicated matters.

Andre

as far as i know, the Lin and Garrelick article is available only by ordering through JASA

it is:

"Sound transmission through periodically frame parallel plates"
JASA, vol 61, #4, pages 1014-1018, 1977


it would be considerably handy to have each scientific discipline organize the various journal articles and research reports into some database.  even if the materails were available only for sale, simply having the capacity to search google-style would be extremely valuable.

Brian

[Brian Dayton]

this is a long term project, ok?  it'll take a while to write all the if's, then's, and formulas but...

i'm so pleased to have found this NRC article yesterday that i will post a spreadsheet who's function is to model the MSM reosonance for each wall type (resilient channel, sound clips, double stud, staggered stud, 2x4 wall, steel stud wall).

what will be given is an estimate of the location of the resonance, and the frequency at which performance surpasses mass law.  it's a real, real long-term project to try to turn this into a prediction sheet, as the formula become exceedingly complex and beyond my capacity to put into excel without reasonable pain.

but it will include the effects of bonding panels, bonding to studs, drywall change, with modified calculations for each wall type.  all based upon experimental data taken here, and cross-checked with the scads of NRC tests (they line up).


if that type of nonsense interests you, stay tuned.

[bert stoltenborg]

Just out of curiousity I looked into the Noise and vibration control engineering from Ver and Beranek, what they have to say about loss of sound in double partitons. I saw that I had been struggling with this years ago (totally forgot, must-drink-less-belgian-beer), and it leads not to convincing results, just to save you the pain in case of...

Bert

[avare]

Thank you Brian!

Verbosely;
Andre

[Dan_Fitzpatrick]

At the risk of becoming the bane of Brian's existence these past few days ... (sorry brian!)

I was just using this very nifty calculator, that Brian created, on my macintosh, and it appears not to work correctly. When in autocalc page or manual calc page, walls do not show a change in MAM frequency (or adjusted cavity volume) depending on whether or not there is insulation specified. this is true for 100% insulation vs. 0% insulation or using any variety of partial fills.

I imagine it probably works fine on PCs, which is what most people are using. I thought it would be good to alert potential mac users to the problem ... unless i am misunderstanding how it is used, which is possible too ... if you like i can provide screenshots showing the behavior.

i wonder then if i am geting no-insulation figures (to which i could multiply .7 ... which would be AWESOME ) or if the numbers i get are completely unusable.

sorry to be a pain Brian

guess i really just need to get a PC

dan

[Dan_Fitzpatrick]

here are some screen shots to show what i mean

the insulation goes from 180mm to 0mm while the cavity depth increases from 50mm to 230mm keeping the net cavity depth the same at 230mm.

yet the Fmam stays the same, as does the adjusted cavity depth (which is supposed to subtract insulation volume when it is present.)

is this a macintosh weirdness or am i doing something wrong?

dan

edit: i have noticed now that the calculations for the Fmam= cell are fairly straightforward ... in fact it appears to be the same "traditional formula" for NO insulation mentioned on page "background" (1900*[(m1+m2)/(d*m1*m2)]^0.5 = Fmam) ...

not sure what is going on, but the sheet does not appear to account for insulation at all in my setup. i think the formula is supposed to automatically change depending on whether there is a number entered for insulation, but it  does not.

hope this helps. in any case, the information in the spreadsheet was very enlightening.

d

[Brian Dayton]

hey dan

the spreadsheet doesn't forbid you from entering insulation that is thicker than the cavity, it just assumes the realization that a 50mm cavity cannot contain 180mm of free insulation.


even if you shoved it in there, it would compress to 50mm, see?  it's not possible


the two calculations that you show screen shots of are as this:

230mm of empty air calculates to 230mm of air (lol)


a 50mm cavity with 180mm of insulation calculates to 230mm, because the formula follows the criteria outlined by the NRC often in the past, which assumes double the effective cavity depth for insulation filled air space.

hence, the net cavity depth can be attained by

actual depth + thickness of insulation


which in this case = 230mm


you cannot put 180mm of insulation in a cavity of less than 180mm, as mentioned above, and the sheet works if this is followed.  (agin, cimpressed insulation doesn't retain it's uncompressed depth)

sorry for the mix-up, and good luck

Brian


PS:  different insulations have different characteristics, and you CANNOT blindly assume that all insulation lowers resonance similarly.  it is far beyond the scope of my lil spreadsheet to attempt to accomodate this, however.

[Dan_Fitzpatrick]

No such thing as a stupid question, right??  :lol:

Thanks for clarifying that, Brian ... and I'm glad it works fine. I didn't know insulation counted for "double," so I learned something new too. Sorry for the trouble!!

Dan  :P

[Brian Dayton]

hey, my bad, one of these days i'll work on making the thing less in need of babysitting to work well


the NRC in some of their papers (IR-586 and many others) and other references have listed 0.71 as the factor by which resonance is lowered when using insulation.  so 100hz to 71hz without and with insulation, respectively.  that corresonds to an equivalent apparent cavity depth of double

USG in an ancient paper presented (as far as i can tell, it's not a very clearly written formula) 1.414 (sqrt of 2) as the factor, not double

but most references deal in doubles, reality deals in another commodity, and you MUST NOT apply that to all wall types